Project developed using Data Mining, Machine Learning and Data Visualization techniques to extract information and make predictions based on positional data.
Thesis work consisting in the acquisition and manipulation of a structured space-time database on which different operations have been performed. The first concerns a graphical analysis of the data using a dashboard; subsequently an algorithm was implemented to perform pattern recognition and finally a recurrent neural network, in particular an LSTM, was implemented to perform prediction operations.
- Programming languages: Python, HTML, SQL
- Libraries: Pandas, Keras, Tensorflow, Plotly, Matplotlib
- Algorithms: PrefixSpan